Until I deleted every last trace of the software after discovering it
was sending IP packets, I had started analysing the Tx sequence to see
if it made any sense

For any given sub-mode, the Tx slots are all equal length, so allocate
a 0 and1 to each Tx state.
If you do an autocorrelation on the resulting sequence, (by XORing a
delayed version of the pattern with itself) some interesting longer
sequences show up - always an even number of 1's' and 0's' for
example.
It almost has a vague Manchester Coding feel to it.

But all analysis is stopped and will permanently stay so until a
version appears that will not send IP packets.
It was much easier to 'crack'  WSJT coding - just a case of reading
the source code :-)

'jnt


On 3 January 2012 17:42, Alberto di Bene <[email protected]> wrote:
> On 1/3/2012 10:36 AM, Andy Talbot wrote:
>
> If 'someone' would publish the innards of that mode, a simple Tx only
> keyer could surely be developed, and te mode woue be useful
>
> And maybe somebody could even adventure in writing an Rx program...
> but of course the protocol needs to be completely open source for this.
>
> 73  Alberto  I2PHD
>
>